The paper presents a family of novel light blob shape descriptors for use in selected active safety algorithms used in Advanced Driver Assistance Systems (ADAS). One of the motivations was to obtain a descriptor that ...
详细信息
Anomaly detection from medical images is badly needed for automated diagnosis. For example, medical images obtained with several modalities, such as magnetic resonance (MR) and confocal microscopy, need to be classifi...
详细信息
This article deals with the issue of identification and quantification of uncertainty components in shielding effectiveness measurements carried out in accordance with the standard EN 50147-1. It describes problems ac...
This article deals with the issue of identification and quantification of uncertainty components in shielding effectiveness measurements carried out in accordance with the standard EN 50147-1. It describes problems accompanying such tasks and presents baseline uncertainty budget. Issues related to uncertainty of the antenna factor, uncertainty of measurement receiver and mismatch of the measurement path were taken into consideration. Further topics requiring consideration have also been identified, such as the type A evaluation of uncertainty, variations of the reflection coefficient at the input of antenna and uncertainties of measurements of S-parameters used to calculate mismatch of the measurement path.
In this paper, a deep learning based object detection model is developed for robotic applications in structured agricultural cultivation of sweet peppers (Capsicum annuum). We propose a realistic simulation based meth...
详细信息
The research presented in this paper focuses on the development of fully automated and autonomous vine suckering, a seasonal viticulture activity required to remove all unwanted shoots from the vine trunks. This requi...
详细信息
The paper presents an analysis of fast selection of neural network for the purpose of visual analysis of mechanical wear on prism lenses of in-pavement airport navigational lighting systems. This issue is particularly...
The paper presents an analysis of fast selection of neural network for the purpose of visual analysis of mechanical wear on prism lenses of in-pavement airport navigational lighting systems. This issue is particularly important in terms of aviation safety and navigational lighting control, regulated by EASA and ICAO. The article is the next stage of the development of the system for the vision control of lamps, in which the concept of using a different neural network with an increased data set prepared by the authors is presented. The Deep Network Designer tool included in the Matlab 2022b environment was used. The solution using the GoogLeNet neural network allows for the classification of lamps with an accuracy of 88.37%.
The paper presents research on the accuracy of measuring illuminance and chromaticity of airport lamps. The impact of the type of DC and AC power supply on measurement was assessed with the use of electronic sensors. ...
The paper presents research on the accuracy of measuring illuminance and chromaticity of airport lamps. The impact of the type of DC and AC power supply on measurement was assessed with the use of electronic sensors. BH1750 and BH1745 type sensors in a microprocessor system with an I 2 C interface were used for the measurements. A professional luxmeter was used for comparison purposes. Experimental tests were carried out under laboratory conditions for three types of halogen lamps: approach system lamps, runway centerline lamps, and touchdown zone lamps.
Steel has emerged as a vital component in various industries, ranging from small tools to large-scale structures like ships and statues. With the increasing demand for steel, the steel manufacturing sector has witness...
详细信息
ISBN:
(数字)9798350381689
ISBN:
(纸本)9798350381696
Steel has emerged as a vital component in various industries, ranging from small tools to large-scale structures like ships and statues. With the increasing demand for steel, the steel manufacturing sector has witnessed significant growth and a corresponding rise in production. However, ensuring high-quality steel has become a paramount concern for producers. Defective steel can have severe consequences during downstream manufacturing processes, leading to accidents and product failures. Consequently, steel industry giants have recognized the need for real-time edge defect detection in steel coils at the strip producing mill, as early detection allows for timely corrective actions. To address this challenge, this research leverages computer vision and deep learning techniques, specifically custom-built defect datasets collected from the industry itself, sourced from AM/NS India. Considering the diverse range of AI-ML methodologies available, the proposed method adopts convolutional neural network (CNN) and advanced CNN-based models such as SSD, Mask RCNN, and Faster RCNN, aiming to identify the most optimal model for building a Real-time Edge Defect Detection System. This research tackles the limitations associated with manual inspection of steel coils by automating the process. By implementing this approach, disruptions in coil transportation can be eliminated, the occurrence of expensive product failures can be minimized, and overall production efficiency can be enhanced. The proposed computer vision-based deep learning approach, once trained and deployed with accurate datasets and dependencies, will introduce a new level of accuracy and advancement in the steel industry's defect detection capabilities.
This paper focuses on bridge inspections performed by intelligent unmanned aerial vehicles (UAV). For this, small data loggers are placed by the UAVs at the bridge, which have to be removed later. In our previous work...
This paper focuses on bridge inspections performed by intelligent unmanned aerial vehicles (UAV). For this, small data loggers are placed by the UAVs at the bridge, which have to be removed later. In our previous work we presented the magnetic localization method SRIOD, which not only detects the center of the data logger but also its orientation. In this work we proof SRIOD in a real application and let a robotic arm locate a magnetic target. The hardware involved is presented including the control strategy for the robot. The motions of the robot and the magnetic target are captured with 12 Optitrack cameras in order to validate the localization quality.
The existence of a minimum measurable length scale was suggested by various theories of quantum gravity, string theory and black hole physics. Motivated by this, we examine a quantum theory exhibiting a minimum measur...
详细信息
暂无评论